The spot-welded joint is widely used as a structural part. Therefore, the nondestructive inspection of the welded region is very important for quality control. In this study, by parallel scanning over the welded region by ultrasonic inspection with a point-focussed probe, its geometry in resistance spot welding of a two-plate joint was inspected. As a result of the inspection, the periphery of the spot weld was identified by the sharp minimum pattern in the scanning graphs of a sound spot weld. when the spot weld included a shrinkage cavity or blow hole in the welded region, different patterns appeared in scanning graphs, so that the tension shearing load determined by testing was also different in the case of these patterns. Therefore, it is probable that the information for evaluating the load is included in the scanning graph obtained by ultrasonic inspection. Our aim is to develop a methodology for nondestructive evaluation of the function of the spot-welded joint. The present study is an attempt to construct an evaluation system for tension shearing load of the joint. The system consists of an ultrasonic inspection device and neural network. The network learned the pattern sets dealing with the interaction between the scanning graphs and the tension shearing load of spot-welded joints. The output from the system was examined by comparing it with the experimental results, and the validity of the system was confirmed.